DeepSeek V4 Pro vs Kimi K3 pricing
- For the Chatbot workload, DeepSeek V4 Pro costs 3.2× less than Kimi K3.
- For the RAG workload, DeepSeek V4 Pro costs 2.5× less than Kimi K3.
- For the Coding agent workload, DeepSeek V4 Pro costs 3.1× less than Kimi K3.
Prices
| USD per 1M tokens | DeepSeek V4 Pro | Kimi K3 |
|---|---|---|
| Input | $1.32 | $3.00 |
| Output | $3.96 | $15.00 |
| Cache read | $0.044 | $0.30 |
| Cache write | Not published | $3.00 |
| Batch input | Not published | Not published |
| Batch output | Not published | Not published |
| Off-peak input | $0.66 | Not published |
| Off-peak output | $1.98 | Not published |
| Off-peak cache read | $0.022 | Not published |
DeepSeek V4 Pro: The Input, Output and Cache read rows are the standard (peak) prices; the Off-peak rows apply at other times. Off-peak is every hour outside 01:00–04:00 and 06:00–10:00 UTC on weekdays, plus weekends and Chinese public holidays.
Verified against the provider's official pricing page on Oct 9, 2026. Official pricing page
Moonshot AI's list price for its own API, as compiled by models.dev and checked Oct 10, 2026, 09:00 UTC. Prices on cloud platforms and resellers can differ. Provider documentation
Monthly cost by workload
| Workload | DeepSeek V4 Pro | Kimi K3 |
|---|---|---|
| Chatbot1,000 in / 500 out tokens, 10,000 requests a day | $990.00/mo | $3,150/mo |
| RAG8,000 in / 500 out tokens, 5,000 requests a day | $1,881/mo | $4,725/mo |
| Coding agent50,000 in / 2,000 out tokens, 1,000 requests a day, 80% cached input | $686.40/mo | $2,160/mo |
Cost = input tokens × input price + output tokens × output price, 30 days a month. Cached input uses the cache read price. Cache write fees are not included.
Context and capabilities
| DeepSeek V4 Pro | Kimi K3 | |
|---|---|---|
| Context | 1,000,000 tokens | 1,048,576 tokens |
| Max output | 393,216 tokens | 1,048,576 tokens |
| Input | text | text, image, video |
| Tool calling | Yes | Yes |
| Reasoning | Yes | Yes |
| Structured outputs | Yes | Yes |
Questions
- What is the price difference between DeepSeek V4 Pro and Kimi K3?
- DeepSeek V4 Pro: $1.32 per 1M input tokens and $3.96 per 1M output tokens. Kimi K3: $3.00 per 1M input tokens and $15.00 per 1M output tokens. For the Chatbot workload (1,000 in / 500 out tokens, 10,000 requests a day), that is about $990.00 per month for DeepSeek V4 Pro and $3,150 for Kimi K3.
- Which has the longer context window, DeepSeek V4 Pro or Kimi K3?
- Kimi K3 has the longer context window: up to 1,048,576 tokens, compared with 1,000,000 for DeepSeek V4 Pro.
- Do DeepSeek V4 Pro and Kimi K3 have prompt caching and batch prices?
- DeepSeek V4 Pro: cached input $0.044 per 1M tokens; no batch price published. Kimi K3: cached input $0.30 per 1M tokens; no batch price published.